Human thyroid tissues collected during surgery were immediately frozen in liquid nitrogen. We examined 13 tumor tissue samples (12 papillary thyroid carcinomas [PTCs] and one Hurthle cell adenoma), and five normal thyroid tissues, which consisted of normal tissues adjacent to the thyroid tumors. All tumors were classified according to the criteria of the World Health Organization Committee. One human follicular cancer cell line (WRO), two human papillary cancer cell lines (TPC1 and K1), and one anaplastic cancer cell line (HTH74) were also investigated. These cell lines were kindly provided by Dr. Adel K. El-Naggar (M.D. Anderson Cancer Center, University of Texas, Houston, TX, USA).

DNA was extracted using standard phenol/chloroform methods. Genomic DNA (2 µg) was denatured in 2 M NaOH for 10 minutes at 37℃. The denatured DNA was diluted in 30 µL freshly prepared solution of 10 mM hydroquinone (Sigma-Aldrich, St. Louis, MO, USA) and 520 µL freshly prepared solution of 3 M sodium bisulfite (Fisher Scientific, Hampton, NH, USA) at pH 5.0, and incubated for 16 hours at 50℃. After incubation, the DNA sample was desalted through a column (Wizard DNA Clean-Up System, Promega, Madison, WI, USA). The DNA was eluted using 50 µL warm water and treated with 5.5 µL 3 M NaOH for 5 minutes at room temperature, and precipitated with ethanol with glycogen as a carrier. The bisulfite-modified genomic DNA was resuspended in 20 µL H₂O and stored at -20℃ until use.

The sodium bisulfite-modified DNA was amplified by PCR. The primers used to amplify RUNX3 were as follows: 5'-TTT GGA GAT ATT TGG GTT TT-3' (sense) and 5'-CCC ATT TAA TAT ACA CAC AAC TAA-3' (antisense). The PCR conditions were as follows: 35 cycles of denaturation at 94℃ for 30 seconds, annealing at 60℃ for 30 seconds, and finally a extension at 72℃ for 30 seconds. PCR products were digested with HpyCH4IV (New England BioLabs, Ipswich, MA, USA). The enzyme-treated DNA products were separated by electrophoresis on 5% PAGE gels. DNA from the blood of normal healthy individuals was used as a negative control. SssI methylase-treated (New England Biolabs) normal lymphocyte DNA was used as a positive control.

Thyroid cancer cell lines were cultured in RPMI1640 supplemented with 10% fetal bovine serum. Two cell lines (WRO and TPC1) were treated with a final concentration of 0.5, 1, and 5.0 µM DAC (Sigma) for 3 days, and/or 300 nM trichostatin A (TSA; Sigma) on the third day. Cells were harvested on the fifth day of treatment and subjected to RNA extraction.

Total RNA was purified from the cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed to single-stranded complementary DNA using random hexamers and MMLV reverse transcriptase. The primers used to amplify RUNX3 were as follows: 5'-ACG CCT ACG TCA TCC TGA AA-3' (sense) and 5'-ATG CCA CAC CTC CTT TCT TA-3' (antisense). Glyceraldehyde 3-phosphate dehydrogenase was used as an internal reference gene. Amplified products were resolved by agarose gel electrophoresis and visualized with ethidium bromide.

COBRA was performed to analyze the methylation status of the RUNX3 CpG islands in human thyroid cancer cell lines and thyroid cancer tissues. Almost all thyroid cancer cell lines showed increased methylation of RUNX3 (Fig. 1). Among the thyroid cancer cell lines studied, WRO and TPC1 had extremely hypermethylated RUNX3, at over 90% (Table 1). RUNX3 was hypermethylated in 10 of the 12 papillary thyroid cancer tissues and was not methylated in five normal thyroid tissues. One Hurthle cell adenoma tissue was not hypermethylated (Fig. 2).

Two human thyroid cancer cell lines (WRO and TPC1) were treated with the demethylating agent DAC and histone-deacetylating agent TSA in an attempt to reactivate RUNX3. DAC treatment resulted in a dose dependent increase in expression of RUNX3 (Fig. 3).